Daily AI news, sourcedUpdates / Research
Research: 222 updates tracked, 28% Official.
222updates tracked
since Sep 28
since Sep 28
By label all items · click to see the list
Most-tracked products & models by Official + Reported items, 2+ only
No product or model has 2+ Official or Reported items yet.
Updates Research · 222 items · newest issue first · page 1 of 12
AreaAllModelsAgents & codingApps & productsChips & infraResearchPricing & limitsBusiness & fundingPolicy & safetyOther
- NVIDIA runs a record 104-qubit fault-tolerant quantum computing simulation on Europe's JUPITER supercomputerOfficialNVIDIA NewsroomOct 20Research
- Zhipu (Z.AI) paper reveals internal unified model GLM-OmniOfficialarXivOct 19Research
- Paper finds recursively self-distilling a safe model on its own safe outputs can make it progressively misalignedOfficialarXiv cs.LGOct 19Research
- Lean4AgenticRL trains 9B and 27B agent models with Lean4-verified process rewards, and the authors report 77.3% on TerminalBench-2.1
- Paper teaches a masked diffusion language model to distrust corrupted context, with a finetune that runs in under five hours on a laptop GPUOfficialarXiv cs.LGOct 19Research
- Paper finds physics-labeled prompts make an LLM a strong early guess for engineering simulator calibration, best used as a warm startOfficialarXiv cs.LGOct 19Research
- ImageEval 2026 entry detects multimodal hallucinations zero-shot with a 7B vision-language model and rotated answer choicesOfficialarXiv cs.CLOct 19Research
- AuditKIT, a source-available evaluation library from Lexsi Labs, puts benchmark, judge, runtime, tool-use and retrieval results in one format
- RealCLI benchmark tests CLI agents against a scripted user who changes instructions mid-task, and the best setup averages 49.2%
- Researchers train GPT-2-small models on 96 synthetic languages to test how language models learn action nominal constructions
- Paper replaces a small language model's learned embedding table with a frozen 16 MB vocabulary and reports matching or better benchmark scoresOfficialarXiv cs.CLOct 19Research
- WarcraftSim, a self-play agent trained on one desktop, beats Warcraft III's built-in insane AI in 89% of gamesUnconfirmedr/MachineLearningOct 19Research
- TempoPFN, a linear-RNN time series forecasting model pre-trained only on synthetic data, is accepted at NeurIPS 2026Unconfirmedr/MachineLearningOct 19Research
- Shanghai AI Lab says its Intern-S2 model helped mathematicians resolve Ghys's conjecture in complex dimension threeOfficialShanghai AI LabOct 17Research
- Google Research, MIT and Harvard study finds AI models leave flaws out of their own work reports unless told to be honest, per The Decoder
- Anthropic says it is hiring a smaller share of junior technical roles as agent hours outnumber employee hours 6 to 1
- Anthropic survey finds its staff expect far faster AI progress and job automation than US adultsOfficialAnthropicOct 17Research
- Anthropic's Frontier Red Team says Claude traced the NIST ECDSA curve seeds to a timestamp-seeded lrand48OfficialAnthropicOct 17Research
- Anthropic says Claude Mythos 5 designed quantum circuits that beat the best published designsOfficialAnthropicOct 17Research
- Stanford, Together AI, UIUC and Meta introduce HybridAgent, agents that think in latent space and act in textOfficialarXivOct 16Research